Welcome to Inside the Source, where private equity, investment banking, and corporate development leaders share inspiring anecdotes from their careers and the insightful advice they’ve learned along the way.
Our very first installment of the series features Michael Gray, Origination Director at Growth Capital Partners (GCP). After graduating with a law degree from the University of Nottingham, Michael found his way into private equity and never looked back. Over the last fifteen years, Michael has managed deal origination at a number of successful UK-based firms, including Lyceum Capital (now Horizon Capital) and Grant Thornton.
As a business development leader, one of his key responsibilities has been helping to oversee these firms’ technology stacks, which is the topic of discussion in this Inside the Source. Read on to learn about which components every firm should consider adding to their origination stack, the biggest technology disappointments over the last several years, how to know when it’s time to retire a tool, and much more.
Sourcescrub: You’ve been in deal origination for nearly two decades now. What are some of the major ways that technology has evolved over that time, and how has that changed and shaped deal origination as a function?
Michael Gray: Fifteen years ago, identifying potential origination targets was done manually through attendee lists from conferences, keyword internet searching, and sector-specific award nomination lists. Once you had built up a list of names and validated scale via Companies House, you had to find phone numbers and email addresses, ultimately picking up the phone to call people, and sending relatively unstructured emails enquiring about an entrepreneur’s ambitions. You would lean on corporate finance advisors for early introductions and ideally some sector insight to help shape a narrative around the reason for your approach.
Today everything is database led, research focused, and thematic in terms of sector mapping and specific fund messaging. The level of data available has materially changed a fund’s ability to identify a business and conduct in-depth research on the company and key individuals prior to engaging. You have database tools like Sourcescrub, Gartner, CapIQ, Mergermarket, etc. that make it so much faster and easier to research entire markets and identify the right targets.
Additionally, we have everything digitally stored. There has been years and years of materials produced, and we are now able, through CRM and our SharePoint folder library, to very quickly work out all the conversations that we have previously had across our core sectors, review any previous analysis done, and shape our messaging with similar targets. We are currently looking into utilizing tools that pull that all into one place.
With it becoming increasingly quick and simple to identify attractive targets and the key individual(s), I can focus much more on quality of conversation, likely issues they’re facing as a business, and where our track record as a fund is potentially relevant, versus getting the conversation itself. Previously, it was all about finding interesting businesses first and then having a conversation with them. Now it’s a case of getting to interesting businesses with a well-thought through, articulate message and having a better-quality conversation than the other parties that a shareholder may already have spoken with.
S: What are some of the tools you have tried and used over the course of your career, and what are some of your favorite tools in GCP’s current tech stack?
MG: At this point I have used many different tools that exist specifically for company research and sourcing, like Gartner, Megabyte, Fame, Beauhurst, Pitchbook, Margermarket, TechMarket Views, Valu8 — there’s an extensive list of available tools which all overlap to one degree or another. I've probably trialed most databases and also used various CRMs including Microsoft Dynamics, Salesforce, Capsule, and Interaction.
Right now at GCP we use Sourcescrub as our primary deal sourcing platform, and it’s really enabled us to take a more specialized approach to our sourcing. Dynamics is our CRM and it serves as our repository of deal-related conversations. It is integrated with Sourcescrub, so any information that Sourcescrub has about a business we’re tracking gets passed to it directly and updated on a regular basis.
S: In your experience, what are the first 2-3 technologies firms should start with when building out a new origination tech stack?
MG: I think the easiest way to determine the answer to this question is to think through what the origination job is. You’re looking to find the names of businesses that roughly fit your investment criteria. Depending on your thesis, you need company specific data highlighting company sector, geography, growth rates, ownership status, etc. And whichever tool gives you the most accurate and relevant information specific to your investment criteria should be your first technology. The fewer platforms you have to touch to get the information you need, the better.
Next, you need a good email client, because you need to contact these people. Usually this is Outlook or Google, and it’s a given. So, really the next technology you need is a CRM platform, because this is where you’re going to keep track of all of your outreach activity. Did your target respond to your email? How long ago did you send it? Should you send a follow-up?
If you’re a larger firm, I would also recommend layering in a tool that surfaces relevant news and specific company updates so that you can layer these into your follow-up emails. And of course, all of these tools should be integrated so that it’s easy to see everything that’s happening in one place.
S: In your opinion, what have been the biggest technology disappointments over the last five years or so?
MG: For me personally, it has been that the technology is still being built by technologists rather than clients. Vendors haven't gone, “What do people live in and use day to day? What are they doing, how are they using it, and how do we best access their way of working?” We're only now just starting to see vendors saying, “Okay, well if clients ultimately live in Google and Outlook, how do we best stitch our product into this?”
Vendors haven't quite worked out how people are actually looking to use the tools. Instead, what they've done is go, “This is the technology. Look at all this clever stuff we can do! By the way, this is how you've now got to behave to get the most out of the tool we’ve built.”
S: If you could build your dream deal origination tech stack, what would that look like and how would it work?
MG: I think technology currently can do a lot more than people have actually worked out applications for. And that's the disconnect. AI is amazing with untapped capabilities, but I don't think companies have yet productized it in ways that people actually want to use it. AI can improve a whole host of processes, and I want to use it in a way that allows me to spend time on more value-add activities. Being able to get a whole host of automation and AI processes in place would enable me to add more business development, sales-focused capability to my team versus having to focus on a data research and analysis skillset because the technology isn’t generating the correct output.
GCP currently invests in businesses that meet specific criteria. We have a track record of over 25 years and we've done 50 investments, so I’m assuming there are some patterns outside of our core investment criteria that we currently have limited visibility of. To the degree it’s possible, I would like an AI technology tool to sit on the top of that extensive library of data and tell us what this unseen criteria is. I would imagine there are ten things that I'm expecting it to identify as I know them already, but it would be invaluable to know what the other ten are.
Once I have identified these unseen criteria, I can start building very specific guidelines for proactive automated searching over databases like Sourcescrub. What I want the tool to do first is rank a list of businesses we have previously identified based on this AI-generated criteria. Once this ranking process has been verified and is accurate, I can turn it into a score rating and apply this logic to larger databases of businesses that have not previously been identified as potential targets. In theory, this could be applied to an impractically large list of businesses across any geography, with appropriate data inputs, and narrowed down quickly to the top 20 and automatically uploaded into our CRM system. This would be Phase One of our AI-led technology upgrade.
The next phase for me is to then support the information gathering phase prior to approaching a business and using Copilot-type technology to templatize and create bespoke, initial outbound emails or call scripts. In theory this will have turned what would have taken three, four, or five hours just to get one email out into a ten-minute task where I'm just validating and personalizing that email, and pressing send. That for me is what good looks like and what we're trying to build towards.
S: What are some of the top things that you look for when trying out a new technology solution and determining whether to add it to your stack?
MG: Ultimately, I look for two things. The first is ease of use — how is clear the UX/UI and how well does it function? The second is the output of the tool. If I’m trying out a data solution, for example, I will ask it something I already know the answer to and see how thorough and accurate the output is. I’ll kick it around for thirty minutes or so, and if it’s able to meet those two criteria, then I will spend more time with and seriously consider it.
S: How do you know when it’s time to retire or replace a tool in your current stack? What are the signs?
MG: Ultimately, you have to look at what people are using the most and what’s adding the most value for your team. There are so many point solutions out there, but is it really worth paying thousands of dollars for something that solves such a small part of the equation when you’re getting more value elsewhere?
It also depends on switching costs. Some tools are a lot harder to replace than others. And last but not least, I would say that the tools that integrate and work well with one another are usually keepers because it’s rare to find tools that communicate well with each other and really streamline operations for you.
S: Are there any new or upcoming technologies that you are particularly interested in and excited about outside of GenAI?
MG: It’s interesting because in my opinion, most tools that claim to be AI have really offered something more akin to automation than genuine AI. To me, true GenAI is much more prescriptive and learns your individual habits and needs over time, it can predict what you want and make tailored recommendations. It still feels a million miles away, but I do think the technology is there. Vendors just need to find a way to productize it. And that’s what I’m truly excited about.
